Research

Musculoskeletal biomechanics (MSB) research team combining experimental, imaging, computational and clinical studies.

 

Imaging of single mineralised collagen fibrils

Postdoc grant UniBE with LBN

Bone is a complex material that offers a unique blend of strength, toughness, and lightness. The key feature contributing to this outstanding combination of material properties is the hierarchical arrangement of bone spanning across the length scale.

TEM image of a mineralised collagen fibre, with a zoom-in view of the mineral crystal.

At the lowest level, bone is built up of mineralized collagen fibrils (MCFs). In this highly collaborative project, we used advanced transmission electron microscopy (TEM) techniques to assess the structural organization of the MCFs. We used TEM to visualize the arrangement of collagen molecules and minerals within the MCFs at the nanoscale. We also measured the orientation of individual mineral crystals within the MCFs. The capabilities of the TEM allow us to resolve MCF organization and composition down to the nanoscale and even Angstrom level, opening-up new venues for research into how bone diseases and treatments affect the ultrastructure of bone.

 

Contribution of Bone Tissue Properties to Strength of the Ageing Human Hip

SNF grant # 200365 with EMPA, PSI, VUT & MUG

To better understand age-related bone fragility, we examine multiscale and multimodal tissue properties in the proximal femur. At the microscale, we define the tissue’s complete mechanical profile (elastic, yield, and post-yield) using nanoindentation and state-of-the-art high-throughput micropillar compression. Enhancing the collected dataset with microradiography measurements provides information on the mechano-mineral interplay. Collaborative synchrotron measurements with PSI enabled us to further investigate the relationship between bone ultrastructure and mechanical properties at the femoral neck, analyzing local variations within bone quadrants and among donors. These experimental approaches were complemented by a computational study at the whole bone level, incorporating strain rate-sensitive material properties into finite element models of the proximal femur. By including strain rate sensitivity in models representing a fall from standing height, the quantitative agreement between simulation and experiment was improved considerably.

 

 

Finite element model of the human proximal femur in a sideways fall configuration and rheological representation of a strain rate-sensitive constitutive model for bone.

 

HR-pQCT-Based Diagnosis of Osteoporosis

with IS & MG

Recently, novel diagnostic tools have been developed to predict the mechanical strength of distal bone sections using homogenized finite elements (hFE) based on high-resolution peripheral computed tomography (HR-pQCT). These tools have lowered repeatability errors for improved longitudinal assessments compared to the gold standard dual X-ray absorptiometry (DXA). The average spatial distribution of bone damage induced by compressive overloading was recently evaluated in a large healthy cohort (n=381, 20-92y) using statistical shape modeling (SSM). Results showed that in the radius damage primarily at the lunate joint within the trabecular compartment, no damage was detected in the cortex. In the tibia, most damage accumulated in the proximal trabecular region, which is characterized by a lower bone volume fraction and a higher cortical thickness. The multi-stack acquisition protocol (20 and 30 mm for radius and tibia, respectively) proved less sensitive to boundary conditions than the standardized single-stack acquisition (10 mm). These findings suggest the potential of hFE-based SSM for understanding bone failure patterns in a population-wide context.

Comparison of the statistical distribution of damage in single versus multiple sections of HR-pQCT-based hFE models of the distal radius and tibia.

 

Improving Primary Stability of Total Hip Arthroplasty

with MUG & indursry partners

Variability of the proximal femoral medullary canal plays a crucial role in cementless femoral stem design and placement in total hip arthroplasty and was quantified using statistical shape modelling based on computed tomography data from 763 femora.

Left: Anterior and superior views of mean medullary canal (blue) and deformations along the first three PCs at ± 2 SD. Right: scree plot for the first 15 PCs.

 

A Fragility Fracture Integrative Risk Model for CT Recycling

SNF grant # 183584 with HUG, IS, & MUG

In the previous year, the development of the required submodels for the novel integrative risk model was completed. The calculator has been implemented as a python pipeline, and the next logic step is its validation using external clinical datasets. Suitable datasets such as the Study of Osteoporosis Fractures (SOF) or the MrOS are currently used to calibrate the model and assess its predictive performance. Preliminary results show that the model performs at least as good as aBMD from DXA alone. Further adjustments to improve the model’s calibration are currently under investigation. Next to that, the follow-up of our own clinical study is in its final stage, and the last follow-up calls are planned for June 2026.

Left: Impact force resulting from a sideways fall. Right: Sensitivity of fracture risk to variations in femoral strength and other anatomical quantities.

 

Biomechanical Stability of Dental and Orthopaedic Implants

Innosuisse grant # 115.975 with ZMK, AO & industry partners

The bone sets the tone for implant stability and surgical success. Finite element simulation combined with CT imaging enables the virtual reconstruction and analysis of a patient’s heterogeneous bone structure, providing insight into the mechanics underlying bone-implant stability. Research in this area advances through experimental investigations and multiscale FE modeling. Studies on carbon fiber-reinforced PEEK pedicle screws revealed their toggling behavior and mechanical performance in spinal fixation. Comparative analyses of micro- and homogenized FE models showed that homogenized simulations can efficiently predict the load-bearing capacity of bone screws. In dental implant surgery, the integration of experimental data with cone beam CT-based FE analyses substantially improved the prediction of primary implant stability in human jawbones. This approach shows potential for clinical implementation, allowing dentists to virtually assess bone quality and optimize implant placement before surgery, thereby improving patient outcomes.

Left: Toggling experiment and FE simulation of a PEEK pedicle screw. Right: Comparison of a microFE with a homogenized FE model.